# Smithery > Add the Spicrawl MCP server from Smithery (spicrawl/web-scraper) with the Smithery CLI or website, send your API key as the x-api-key header, and use all 24 spicrawl_* tools. Source: https://docs.spicrawl.com/agents/smithery [Smithery](https://smithery.ai) is a registry and gateway for MCP servers: you pick a server on its website or CLI, and Smithery connects your AI client to it. Spicrawl is listed there as [`spicrawl/web-scraper`](https://smithery.ai/servers/spicrawl/web-scraper). The listing proxies to the hosted Spicrawl MCP server (`https://mcp.spicrawl.com/mcp`), so every tool runs under your API key with the same scopes, limits and credits as a direct API call. Install it into an MCP client with one command (restart the client afterwards): ```bash npx -y smithery mcp add spicrawl/web-scraper --client cursor ``` ## Add Spicrawl from Smithery **Smithery CLI** The [Smithery CLI](https://smithery.ai/docs/concepts/cli) needs Node.js 20 or later. `--client` writes the server into that client's own config. The accepted names include `claude-code`, `claude`, `cursor`, `windsurf`, `vscode`, `gemini-cli`, `codex`, `cline`, `opencode` and `goose`; `npx -y smithery mcp add --help` lists them all. ```bash npx -y smithery mcp add spicrawl/web-scraper --client claude-code ``` This registers `https://server.smithery.ai/spicrawl/web-scraper/mcp` in the client. The CLI reports the transport as `http-oauth`: the client authenticates to Smithery, and no Spicrawl key is stored. Restart the client, then check that the 24 tools are listed. To keep the connection in your Smithery account and send your Spicrawl key with it, sign in and pass the key as a header. Smithery stores headers securely and never reads them back: ```bash npx -y smithery auth login npx -y smithery mcp add spicrawl/web-scraper --id spicrawl \ --headers "{\"x-api-key\": \"$SPICRAWL_API_KEY\"}" npx -y smithery tool list spicrawl ``` Then call a tool to check the key works. This one lists your latest API call and spends no credits: ```bash npx -y smithery tool call spicrawl spicrawl_requests_list '{"limit": 1}' ``` **Smithery website** **Step 1: Open the listing** Go to [smithery.ai/servers/spicrawl/web-scraper](https://smithery.ai/servers/spicrawl/web-scraper). The page lists the 24 tools and their descriptions. **Step 2: Add it to your toolbox** Select **Add to toolbox**, then **Sign in to install** and sign in to Smithery. **Try now** opens a menu of chat apps to install it into (Telegram and Slack when this page was written). **Step 3: Enter your API key** The connection has one setting, **Spicrawl API key**. Paste a key from [app.spicrawl.com](https://app.spicrawl.com) (`spicrawl_live_...`). Smithery sends it to Spicrawl as the `x-api-key` header, never in the URL. You can leave it empty to browse the tools first (see below). ## API key The key is optional when you connect and required when you call a tool. | | Without a key | With a key | | ----------------------------- | --------------------- | --------------------------------------------- | | Connect and list the 24 tools | Works | Works | | Call any tool | Fails with HTTP `401` | Runs under the key | | Credits | None charged | Successful calls are charged; failures cost 0 | Without a key, a tool call returns this error, which names the header to fix: ```text Unauthorized: send your Spicrawl API key as 'Authorization: Bearer spicrawl_live_...' (or in an 'x-api-key' header) ``` Create a key in [app.spicrawl.com](https://app.spicrawl.com). Keys start with `spicrawl_live_`. Through Smithery the key travels in the `x-api-key` header, so it never appears in a URL or in a shared config file. A new key has the `scrape`, `batch` and `sessions` scopes; the usage tools need `read`. See [Authentication](https://docs.spicrawl.com/authentication.md) and the [MCP server](https://docs.spicrawl.com/agents/mcp.md#authentication) page. ## Tools Smithery lists 24 tools, all named `spicrawl_*`. The full argument reference for each is on [MCP server](https://docs.spicrawl.com/agents/mcp.md#tools). | Group | Tools | What for | | ----------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------- | | Scrape | `spicrawl_scrape` | One URL as markdown, text, HTML, JSON or PDF, with optional extraction, links and screenshots. | | Batch | `spicrawl_batch_submit`, `spicrawl_batch_list`, `spicrawl_batch_status`, `spicrawl_batch_results`, `spicrawl_batch_task_content`, `spicrawl_batch_cancel`, `spicrawl_batch_retry`, `spicrawl_batch_add_items`, `spicrawl_batch_close` | Tens to thousands of URLs as one asynchronous job. | | Sessions | `spicrawl_session_create`, `spicrawl_session_list`, `spicrawl_session_get`, `spicrawl_session_context`, `spicrawl_session_release`, `spicrawl_session_delete` | A persistent identity (cookies, storage) for multi-step flows such as logging in. | | Usage and history | `spicrawl_usage`, `spicrawl_usage_summary`, `spicrawl_usage_reconciliation`, `spicrawl_requests_list`, `spicrawl_request_get` | Credits used and recent API calls. The three `spicrawl_usage*` tools need the `read` scope. | | Docs | `spicrawl_docs_search`, `spicrawl_docs_read`, `spicrawl_docs_index` | Let the agent look up a parameter or an error code in these docs. | `spicrawl_browser_connect_url` (coming soon) is not listed on Smithery. Since 0.1.7 each tool also declares an MCP `outputSchema` and returns `structuredContent` beside the text, so a client can read fields such as `credits` and `status` without parsing text. See [Structured results](https://docs.spicrawl.com/agents/mcp.md#structured-results). ## Troubleshooting | Symptom | Cause | Fix | | ------------------------------------------------------------------------------------------------ | -------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | A tool call returns HTTP `401` and `Unauthorized: send your Spicrawl API key ...` | No Spicrawl key reached the server. | Enter the key in the connection's **Spicrawl API key** setting, or pass `--headers` with `x-api-key` as above. Create a key in [app.spicrawl.com](https://app.spicrawl.com). | | A tool call still returns `401` after you added a key | The key is unknown, revoked or expired, or was pasted with extra characters. | Copy the key again, or create a new one in the dashboard. | | `curl https://web-scraper--spicrawl.run.tools` returns `401` with `Missing Authorization header` | That address is Smithery's gateway. It wants Smithery credentials, not a Spicrawl key. | Connect through Smithery (CLI or toolbox). To use your Spicrawl key directly, connect to `https://mcp.spicrawl.com/mcp` instead ([MCP server](https://docs.spicrawl.com/agents/mcp.md)). | | `ERR::AUTH::INSUFFICIENT_SCOPE` (HTTP 403) from a `spicrawl_usage*` tool | The key lacks the `read` scope. | Grant `read` to the key in the dashboard. | | `ERR::LIMIT::QUOTA_EXCEEDED` (HTTP 402) | Out of credits or over the monthly ceiling. | Not retryable. The error names the reset date; see [Credits](https://docs.spicrawl.com/credits.md). Failed calls cost 0 credits. | | The tools do not appear in the client | The client was not restarted after the install. | Restart the client and check its MCP panel. | ## Links * Listing: [smithery.ai/servers/spicrawl/web-scraper](https://smithery.ai/servers/spicrawl/web-scraper) * Smithery docs: [smithery.ai/docs](https://smithery.ai/docs) and the [CLI reference](https://smithery.ai/docs/concepts/cli) * Spicrawl MCP server and tool reference: [MCP server](https://docs.spicrawl.com/agents/mcp.md) * Other directories and marketplaces: [Where to find Spicrawl](https://docs.spicrawl.com/agents/integrations.md)